`hls-hlint-plugin` test failure in Gentoo Linux

[hololeap]

I'm getting a test failure for hls-hlint-plugin-1.1.0.0:

    .hlint.yaml fixity rules are applied:                                                                                     FAIL (8.52s)
      src/Test/Hls/Util.hs:308:
      Got unexpected diagnostics for Uri {getUri = "file:///var/tmp/portage/dev-haskell/hls-hlint-plugin-1.1.0.0/work/hls-hlint-plugin-1.1.0.0/test/testdata/fixity/FixityUse.hs"} got [Diagnostic {_range = Range {_start = Position {_line = 5, _character = 13}, _end = Position {_line = 5, _character = 46}}, _severity = Just DsInfo, _code = Just (InR "refact:Functor law"), _source = Just "hlint", _message = "Functor law\nFound:\n  show <$> mInt <!> pure <$> Just c\nWhy not:\n  show . (mInt <!> pure) <$> Just c\n", _tags = Nothing, _relatedInformation = Nothing}]
      Use -p '/.hlint.yaml fixity rules are applied/' to rerun this test only.

---

Here is the diagnostic message pretty-printed:

[ Diagnostic
    { _range = Range
        { _start = Position
            { _line = 5
            , _character = 13
            }
        , _end = Position
            { _line = 5
            , _character = 46
            }
        }
    , _severity = Just DsInfo
    , _code = Just
        ( InR "refact:Functor law" )
    , _source = Just "hlint"
    , _message = "Functor law
      Found:
        show <$> mInt <!> pure <$> Just c
      Why not:
        show . (mInt <!> pure) <$> Just c
      "
    , _tags = Nothing
    , _relatedInformation = Nothing
    }
]

---

Full output of the following command is here:

$ LSP_TEST_LOG_STDERR=1 LSP_TEST_LOG_MESSAGES=1 ./setup test --show-details=streaming --test-option=--color=always --test-option=+RTS --test-option=-T

---

Your environment

• GHC 9.0.2 on Gentoo Linux. All packages are installed through the system package manager.
• HLint v3.4.1

This seems to be unrelated to my other issue, as I am getting this failure when I run the tests outside Gentoo's sandbox, so it seems to be unrelated to any sandboxing issues.

[michaelpj]

Are you using the same index-state as HLS? Could you be getting a newer version of hlint than we're testing with?

[hololeap]

We're using hlint-3.4.1 and it looks like your tests are being run with hlint-3.2.8 from what I can see.

[michaelpj]

Okay, so newer versions of hlint might add more hints, so this is not necessarily an unexpected failure. It would be ideal if our tests were robust to that, but I have no idea if they are.